Formalising common approaches to trading
This thread was created to systematise approaches to trading. I'll post my thoughts on it to start with so I don't forget :)
What is needed to gain an advantage? Based on the organisation of bidding (double auction) the forces are four. Quoted markets also essentially boil down to this pattern. The first two represent buy and sell liquidity (limit orders). The second two buy on the market and sell on the market (stop orders are essentially also on the market). Those for the market create the movements and in fact pay for them out of their own pockets (slippage). Limit ones counteract those movements. Dense Limit Sheet Areas create resistance and support levels/zones. Zones/zones of market surge and stop orders create price drive zones. These zones are unevenly distributed in price and time.
The task of trading is to learn at certain moments to correctly assess the ratio of potentials of these forces (not necessarily all four). Not only their current distribution, but also for some time ahead. For example, the same sanction is an attempt to estimate the accumulation of potential buyers or sellers in the market. The potential is high and probably it will be realized, which will lead to the price drive in the corresponding direction. How to estimate this potential - there are several options. But all of them rest on peculiarities of trading, typical behaviour of traders, limited capital at their disposal, etc. All of these can be called constraints.
The task of trading is to learn at certain moments to correctly assess the ratio of potentials of these forces (not necessarily all four). Not only their current distribution, but also for some time ahead. For example, the same sanction is an attempt to estimate the accumulation of potential buyers or sellers in the market. The potential is high and probably it will be realized, which will lead to the price drive in the corresponding direction. How to estimate this potential - there are several options. But all of them rest on peculiarities of trading, typical behaviour of traders, limited capital at their disposal, etc. All of these can be called constraints.
In general, there are 4 forces: liquidity to sell, liquidity to buy, market buyers, market sellers. You can abbreviate: LS, LB, MB, MS.
You can represent them as functions of price and time, and the output is volume to buy or sell: LS(P,t)=V, etc.
It is clear that it is impossible to determine the exact values of these functions for future moments of time and price. In fact, it is not necessary - their ratios and not at the individual price/time points, but the sum over some range of price/time are enough.
Ratios like MB(dp1,dt1)>>MS(dp1,dt1) buyers in the market at a certain time interval and prices much more sellers. And if also MB(dp1,dt1)>LS(dp1,dt1) then the price will rise by more than dp1 in time dt1.
Let me say right away that I do not have a strict formalization of this mathematically, as well as the need for it. It is simply for convenience of perception. Certain ratios of these forces lead to certain events that are reflected on the chart. If, for example, liquidity is low at some point of time in the traded range relative to the market side, the volatility will be high, and the price increment in the end will depend on the ratio of market orders to buy and sell.
The important thing is of course that these forces are interdependent on previous values. To put it simply, market buyers who open positions at some time (or at some prices) will become market sellers or liquidity sellers. Basically, these relationships are tracked on the history to predict with some probability the correlation of these forces in the future.
You can represent them as functions of price and time, and the output is volume to buy or sell: LS(P,t)=V, etc.
It is clear that it is impossible to determine the exact values of these functions for future moments of time and price. In fact, it is not necessary - their ratios and not at the individual price/time points, but the sum over some range of price/time are enough.
Ratios like MB(dp1,dt1)>>MS(dp1,dt1) buyers in the market at a certain time interval and prices much more sellers. And if also MB(dp1,dt1)>LS(dp1,dt1) then the price will rise by more than dp1 in time dt1.
Let me say right away that I do not have a strict formalization of this mathematically, as well as the need for it. It is simply for convenience of perception. Certain ratios of these forces lead to certain events that are reflected on the chart. If, for example, liquidity is low at some point of time in the traded range relative to the market side, the volatility will be high, and the price increment in the end will depend on the ratio of market orders to buy and sell.
The important thing is of course that these forces are interdependent on previous values. To put it simply, market buyers who open positions at some time (or at some prices) will become market sellers or liquidity sellers. Basically, these relationships are tracked on the history to predict with some probability the correlation of these forces in the future.
For me, any inefficiencies are based on real limitations and discreteness of a particular market. For example, the separation of traders by position holding horizon, or by placing stop orders, liquidity allocation, etc. Purely technical features associated with the trading organization, stereotypes in trading, time, legislative, financial constraints.
This is the division of traders into horizons of position holding and their stereotypical behavior that determines the market memory. The horizon itself forms "account memory" - i.e. while holding a position and having an unrecorded paper profit or loss, they are forced to remember, and the depth of memory is determined by the time the position was held. The second, which follows in part from the first, is the average depth of history analysis for traders (technical analysis) of that horizon. Also, non-speculative traders can contribute to the memory related to their trading cycle. For example, the hedgers dealing with agricultural products can have their cycle determined by the crop cycle etc. In general, traders can be divided into groups/subgroups with similar cyclicality in trading (as a rule, not periodical). This cyclicality is what determines market memory. Of course, this division is relative and there are many ways to divide traders including the meaningless ones. The sensible group is the one whose participants start and/or finish their trading cycle (opening a trade - closing it) at almost the same moments. They walk in "friendly rows", as Neo wrote. Basically, the correct division into these groups and the way to calculate the beginning and the possible end of their cycle on the price series is the basis to predict the future correlation of forces in the bulls/bears market/limit. It is the increasing necessity for the group to exit the trade. Why, wrote above. The internal time is determined again within the framework of a particular group and its cycle. Practically, the task is to calculate the cumulative pose of the group under observation on the time series of the beginning of the cycle. But this is only one method, usually in a counter-truth implementation. What Neo or FelixWhite wrote about. Although I must say right away that the latter has not written anything specific and I cannot judge his successes. I mentioned him because he is often recalled))).
P.S. Within the framework of limit/market bulls/bears it is convenient to think and set goals for other trading methods. And not every method needs a group allocation for example. There are simpler methods as well.
This is the division of traders into horizons of position holding and their stereotypical behavior that determines the market memory. The horizon itself forms "account memory" - i.e. while holding a position and having an unrecorded paper profit or loss, they are forced to remember, and the depth of memory is determined by the time the position was held. The second, which follows in part from the first, is the average depth of history analysis for traders (technical analysis) of that horizon. Also, non-speculative traders can contribute to the memory related to their trading cycle. For example, the hedgers dealing with agricultural products can have their cycle determined by the crop cycle etc. In general, traders can be divided into groups/subgroups with similar cyclicality in trading (as a rule, not periodical). This cyclicality is what determines market memory. Of course, this division is relative and there are many ways to divide traders including the meaningless ones. The sensible group is the one whose participants start and/or finish their trading cycle (opening a trade - closing it) at almost the same moments. They walk in "friendly rows", as Neo wrote. Basically, the correct division into these groups and the way to calculate the beginning and the possible end of their cycle on the price series is the basis to predict the future correlation of forces in the bulls/bears market/limit. It is the increasing necessity for the group to exit the trade. Why, wrote above. The internal time is determined again within the framework of a particular group and its cycle. Practically, the task is to calculate the cumulative pose of the group under observation on the time series of the beginning of the cycle. But this is only one method, usually in a counter-truth implementation. What Neo or FelixWhite wrote about. Although I must say right away that the latter has not written anything specific and I cannot judge his successes. I mentioned him because he is often recalled))).
P.S. Within the framework of limit/market bulls/bears it is convenient to think and set goals for other trading methods. And not every method needs a group allocation for example. There are simpler methods as well.
When applied to critical sentiment accumulation methods, the problem statement looks something like this:
1. define a target group that walks amicably. More precisely, what can be a link for such a group. The simplest one is the horizon of holding the pose. For example for intraday players it is quite clear what the boundaries of their cycle are.
2) Determine based on 1 the starting points of the accumulation cycle. Perhaps in some cases 1 of the beginning of the cycle can be considered the end of the previous cycle. Then we need a formal method of finding the end of the previous one.
3. A method of calculating the accumulation process of an aggregate position/sentiment of this group from the price series. It means a certain elementary price action which increases the considered product range by a certain calculated value. There can be several types of price actions and the increase in the price range is rather arbitrary. Ideally, the price action and the calculation method should logically derive from 1.
4. The way of verification of the resulting model. I think it is a trading system or several variants implementing the final stage of the accumulation process - closing of positions.
1. define a target group that walks amicably. More precisely, what can be a link for such a group. The simplest one is the horizon of holding the pose. For example for intraday players it is quite clear what the boundaries of their cycle are.
2) Determine based on 1 the starting points of the accumulation cycle. Perhaps in some cases 1 of the beginning of the cycle can be considered the end of the previous cycle. Then we need a formal method of finding the end of the previous one.
3. A method of calculating the accumulation process of an aggregate position/sentiment of this group from the price series. It means a certain elementary price action which increases the considered product range by a certain calculated value. There can be several types of price actions and the increase in the price range is rather arbitrary. Ideally, the price action and the calculation method should logically derive from 1.
4. The way of verification of the resulting model. I think it is a trading system or several variants implementing the final stage of the accumulation process - closing of positions.
I will try to set a general speculative trading objective.
1. You need movement in order to make a profit. That is, the market side must be stronger than the corresponding limit side. For example, to buy, it is necessary to enter before the moment when the MB side absorbs the LS side within a significant price range. I.e. it is necessary to determine the moment of a certain drive.
2. You should enter before or at the beginning of the drive. For the same buy, you can enter on the MB side or on the LB side. The first option is to enter at the start of the target drive. The second is an additional prediction that LB>MS.
3. you need to exit at some stage of the drive in time. You can exit by market or by limits, i.e. on the MS or LS side.
In view of the above, the overall challenge is point 1. What is needed is a way to identify moments of increased activity by buyers or sellers. This activity can be caused by external factors, such as the influx of new money, and internal factors, caused by the behaviour of the price and the reaction of participants to it. The methods using the first factors include trend following. The second are more interesting. We should pay more attention to these factors. And most importantly, the methods of their timely identification.
a) Sanitation methods. We can distinguish the exit sanitation - this is when buyers or sellers close their trades. Entry sanction - general methods of entering by TA, or other reasons for entry, but calculated through price/time/volume.
This is the most extensive section. Based on the fact that the publicly available source of information for traders is price, time and volume, the drive points are defined through them or their combinations. The simplest methods are using only one component of the three. For example, the price. Traders unequally place orders to open and close a trade by price. In its pure form it is round levels and extremums.
The second is the astronomical time separately. This is primarily the time of day and news releases. Especially for Forex, where it by itself determines the balance of power between the limit and the market side. Which uses the same night systems, different effects on the session boundaries, etc. Using volume in isolation from price and time doesn't seem to have any effects.
But most methods use combinations of time price and volume to identify drive moments. Which combination depends on which combination most use (financially speaking). You can start with elementary pair combinations. The simpler the better to some extent.
The combination of price and time. The simplest one is to change the price in a fixed time. The second one is to additionally fix the beginning of counting by the astronomical time (or range of time). For example, this is changing of price from the beginning of some session or after the news release. The third way is taking some price as a reference point - the price change for a fixed time after the price has passed some level, for example. Also, the purpose of the calculation may be not the price or its change, but the time change. For example, time of reaching a new extreme, time of price being at a certain level(Market Profile).
Interesting variants are also provided by combinations of price and volume and time and volume.
I think that more complex variants that use all three components are more correct and convenient to calculate through a combination of simpler ones (in pairs and separately). The more complex the combination, the less chance it has of spreading, and hence of correctly reflecting others' intentions. imha. The effectiveness of certain combinations depends on the microstructure of the particular market. Mainly it is the composition of the participants and the rules of the organisation of trading.
b) Cyclical change in activity - buyers' activity in the market is replaced by sellers' activity. Swing trading seems to be at the heart of it
1. You need movement in order to make a profit. That is, the market side must be stronger than the corresponding limit side. For example, to buy, it is necessary to enter before the moment when the MB side absorbs the LS side within a significant price range. I.e. it is necessary to determine the moment of a certain drive.
2. You should enter before or at the beginning of the drive. For the same buy, you can enter on the MB side or on the LB side. The first option is to enter at the start of the target drive. The second is an additional prediction that LB>MS.
3. you need to exit at some stage of the drive in time. You can exit by market or by limits, i.e. on the MS or LS side.
In view of the above, the overall challenge is point 1. What is needed is a way to identify moments of increased activity by buyers or sellers. This activity can be caused by external factors, such as the influx of new money, and internal factors, caused by the behaviour of the price and the reaction of participants to it. The methods using the first factors include trend following. The second are more interesting. We should pay more attention to these factors. And most importantly, the methods of their timely identification.
a) Sanitation methods. We can distinguish the exit sanitation - this is when buyers or sellers close their trades. Entry sanction - general methods of entering by TA, or other reasons for entry, but calculated through price/time/volume.
This is the most extensive section. Based on the fact that the publicly available source of information for traders is price, time and volume, the drive points are defined through them or their combinations. The simplest methods are using only one component of the three. For example, the price. Traders unequally place orders to open and close a trade by price. In its pure form it is round levels and extremums.
The second is the astronomical time separately. This is primarily the time of day and news releases. Especially for Forex, where it by itself determines the balance of power between the limit and the market side. Which uses the same night systems, different effects on the session boundaries, etc. Using volume in isolation from price and time doesn't seem to have any effects.
But most methods use combinations of time price and volume to identify drive moments. Which combination depends on which combination most use (financially speaking). You can start with elementary pair combinations. The simpler the better to some extent.
The combination of price and time. The simplest one is to change the price in a fixed time. The second one is to additionally fix the beginning of counting by the astronomical time (or range of time). For example, this is changing of price from the beginning of some session or after the news release. The third way is taking some price as a reference point - the price change for a fixed time after the price has passed some level, for example. Also, the purpose of the calculation may be not the price or its change, but the time change. For example, time of reaching a new extreme, time of price being at a certain level(Market Profile).
Interesting variants are also provided by combinations of price and volume and time and volume.
I think that more complex variants that use all three components are more correct and convenient to calculate through a combination of simpler ones (in pairs and separately). The more complex the combination, the less chance it has of spreading, and hence of correctly reflecting others' intentions. imha. The effectiveness of certain combinations depends on the microstructure of the particular market. Mainly it is the composition of the participants and the rules of the organisation of trading.
b) Cyclical change in activity - buyers' activity in the market is replaced by sellers' activity. Swing trading seems to be at the heart of it
here's an example of how you can try to single out traders of a particular horizon based on the price/volume combination. We look at the swings of a certain parameter. We have, for example, an upward (impulse) swing and a correctional swing after it. We assume that the corrective swing is mainly caused by profit taking by the participants of the lower TF. Calculate the total volumetric volume of the impulse and correction swing. Subtract the correctional volume from the impulse volume. The received value is an approximate estimation of the position volume of players with higher TF (and of price moves respectively). Then we can roughly estimate their average entry level, e.g. by the median of volum distribution according to the impulse swing, or in a more trivial way. It can be considered as a new reference point in terms of both price and time. A move towards the impulse swing means a paper profit for the group in question, a move below that means a loss. If the next swing is again an impulse and a correction, then we will have 2 reference points and the total volumetric value of the positions of the traded players of the group in question. The criticality of the accumulated volum, as well as the price and time objectives of the players, determine the points of entry - they will create the drive.
Of course, this is also some averaging/coining. Can't do without that. As I wrote above, there should be fairly simple models, but at the same time reflecting what we are looking for accurately enough. And then practice is the criterion of truth - testing of specific strategies based on hypotheses, analysis of results, additional research, new hypotheses, etc. After all, we don't know the microstructure of a specific market a priori - only hypotheses and their testing. Perhaps even simpler hypotheses will be more effective. For example, consider candlesticks of a certain frame without swings. For example, to consider as a correction the upper shadow of a bullish candlestick (the lower shadow of a bearish candlestick).
Of course, this is also some averaging/coining. Can't do without that. As I wrote above, there should be fairly simple models, but at the same time reflecting what we are looking for accurately enough. And then practice is the criterion of truth - testing of specific strategies based on hypotheses, analysis of results, additional research, new hypotheses, etc. After all, we don't know the microstructure of a specific market a priori - only hypotheses and their testing. Perhaps even simpler hypotheses will be more effective. For example, consider candlesticks of a certain frame without swings. For example, to consider as a correction the upper shadow of a bullish candlestick (the lower shadow of a bearish candlestick).
Thanks for the insightfulness, very interesting.
О! And no battles with aliens).
For myself, if you poke around some more, it turns out that analysing one instrument is like analysing the movement of one car when studying the whole city flow.
The market has linked everything and currencies and energy and metals and food and stocks. If you go on and on, personally I suddenly found myself at the interest rates, though
even though it's perfectly logical and natural.
sever31:
Thanks for the insightfulness, very interesting.
Thanks for the insightfulness, very interesting.
Please, I hope we can develop the topic together
Mischek:
О! And no battles with aliens.)
For myself, if you poke around some more, it turns out that analysing one instrument is like analysing the movement of one car when studying the whole city flow.
The market has linked everything and currencies and energy and metals and food and stocks. If you go on and on, personally I suddenly found myself at the interest rates, though
even though it's perfectly logical and natural.
yes, interconnecting markets is an interesting topic and a very influential part of the market trading different spread trading variants. Much of the volume is generated by them
It is interesting to ponder what underlies the links between markets and instruments. Probably, in the beginning it was a connection through the common production or financial cycle, but now this connection is mainly in the heads of investors and not a rational orientation to some fundamentals. There is an orientation to them, but through the prism of stereotypes existing in the market now. That is, it is based on methods of calculating the over/undervaluation of an asset based on comparison with the prices of other assets, or some combination of assets. Investors look for undervalued assets compared to peers (as they think). Speculators pick up on these trends and try to get in and out before investors, reducing their potential gains or even increasing their losses. Thereby reinforcing the connection.
Another question is whether one should try to profit from investors by becoming a spreader or even from the mistakes of speculators and how it can be done. Or how to "clear" prices from spread traders' trades. There's a lot to discuss.))
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register